DescriptionOn-chip power-delivery-network (PDN) is an essential element of physical implementation that strongly determines quality and reliability. During signoff, multiple ECO iterations are needed to ensure that each instance of the design should meet IR drop specification. Even though the design remains very similar after ECO changes, conventional PDN analysis in industry standard CAD tool takes longer runtime to determine IR drop. This leads to a much longer PDN analysis time for each ECO iteration. Our goal is to reduce this runtime in each iteration to evaluate the ECO changes and fix the violating cells immediately, prior to run conventional signoff tool. Hence improving the number of iteration and achieving faster convergence. In this paper, we have developed a Machine Learning (ML) framework to address this problem. Based on ML prediction, we first address the violating cells until all of them meet the IR drop specification. This process may take many iterations, but as the runtime of each iteration is only few minutes, it is obvious to achieve a faster IR drop convergence. Next, we run industry standard tool for detail simulation to see if any outlier still exists. Hence, minimum number of iterations is expected in IR drop simulation tool.